A system that replaces whole words or phrases with arbitrary symbols (usually numbers). Unlike a cipher, which transforms individual letters according to a rule, a code has no mathematical relationship between the original word and its replacement. "ATTACK" might be code 7392 while "RETREAT" is 5023 — there is no pattern you can use to predict one from the other.

Used on: Codebooks, Zimmermann Telegram, Black Chamber, CODEX Codebook Suite

A book (or document) that lists the correspondence between words/phrases and their code numbers. Both the sender and receiver must have a copy. A codebook typically has two sections: an encoding section (sorted by word) for the sender, and a decoding section (sorted by number) for the receiver. Diplomatic codebooks could contain tens of thousands of entries.

Used on: Codebooks, Zimmermann Telegram, Black Chamber, CODEX Codebook Suite

A single unit in a coded message — typically a four- or five-digit number that represents one word or phrase from the codebook. The Zimmermann Telegram, for example, was transmitted as a series of code groups like "13042 13401 8501."

Used on: Codebooks, Zimmermann Telegram, CODEX Codebook Suite

A hybrid system combining a small codebook (for important words and names) with a substitution cipher (for everything else). Nomenclators were the standard tool of European diplomatic communication from the 1400s through the 1800s. They were more secure than pure ciphers, but the cipher portion was still vulnerable to frequency analysis.

Used on: Codebooks

In a codebook or cipher, a homophone is one of several different code groups that all mean the same thing. For example, a codebook might give the common word “the” twelve different codes, picked at random each time it appears. This is called homophonic substitution. It is the single most effective defense against simple frequency analysis: because the most common plaintext words now produce many different code groups instead of just one, no code group stands out as obviously meaning “the” or “and.” The defender pays for this in book size — the codebook has to be much larger — and the attacker responds with techniques like distributional clustering that try to group the homophones back together.

Used on: CODEX Codebook Suite

A meaningless symbol inserted into a coded or enciphered message to confuse cryptanalysts. A null adds no information to the message — the receiver simply ignores it. Nulls make it harder to analyze the real content by disrupting patterns and inflating the apparent number of different code groups.

Used on: Codebooks, CODEX Codebook Suite

A second layer of encryption applied on top of a code. After encoding a message with a codebook, the code numbers are then modified using a cipher (for example, adding a secret number sequence to each code group). This means that even if an enemy has captured the codebook, they still cannot read the message without also knowing the superencipherment key.

Used on: Codebooks, Zimmermann Telegram, CODEX Codebook Suite

A coded message that has not been superenciphered — the code groups are transmitted exactly as they appear in the codebook. A plaincode message is easier to break than a superenciphered one, because the same word always produces the same code number, allowing pattern analysis over time.

Used on: Codebooks

The study and practice of breaking codes and ciphers — reading secret messages without having the key. Cryptanalysis encompasses everything from simple frequency analysis to sophisticated statistical methods like the Index of Coincidence. William Friedman transformed it from a craft into a branch of applied mathematics.

Used on: Early Modern Cryptanalysis, Codebooks, Zimmermann Telegram, Black Chamber, Friedman

A person who practices cryptanalysis — someone who breaks codes and ciphers. In the early 20th century, cryptanalysts were often linguists, classicists, or mathematicians recruited for their pattern-recognition skills. Friedman's innovation was to make cryptanalysis more systematic and less dependent on individual brilliance.

Used on: Codebooks, Zimmermann Telegram, Black Chamber, Friedman

A statistical measure developed by William Friedman that calculates the probability that two randomly chosen letters from a text are the same. English text has an IC of about 0.0667 (because common letters like E and T appear often). Random text has an IC of about 0.0385 (exactly 1/26). The IC can reveal what type of cipher was used: a monoalphabetic cipher preserves the English IC, while a polyalphabetic cipher pushes it toward the random value. The IC can also be used to estimate key length.

Used on: Friedman

A method for finding the key length of a polyalphabetic cipher (like Vigenère). It works by finding repeated sequences of letters in the ciphertext. The distance between repetitions is always a multiple of the key length. By finding the greatest common divisor (GCD) of all these distances, you can determine the most likely key length. Named after Friedrich Kasiski, who published the method in 1863.

Used on: Friedman, CODEX Codebook Suite

A statistical test that measures how well an observed distribution matches an expected distribution. In cryptanalysis, it is used to compare the letter frequencies in a ciphertext to the expected frequencies of English. A low χ² value means the text closely matches English; a high value means it does not. This is useful for breaking Caesar ciphers: try all 26 shifts and pick the one with the lowest χ².

Used on: Friedman

The first American military textbook on cryptanalysis, written by Parker Hitt and published in 1916. It covered practical techniques for breaking substitution and transposition ciphers, including frequency analysis, digraph and trigraph analysis, and methods for attacking polyalphabetic ciphers like Vigenère. The Manual trained a generation of American military cryptanalysts and was among the texts William Friedman studied as he developed his own methods.

Used on: Hitt, Friedman, and Scientific Cryptanalysis, Early Modern Cryptanalysis

A cipher device consisting of a set of alphabet strips — each with the 26 letters in a different scrambled order — mounted on a cylinder or frame. To encrypt, the user arranges strips so one row spells out the plaintext, then reads off a different row as ciphertext. Parker Hitt designed the first American strip cipher around 1914. His design evolved into the U.S. Army's M-138 strip cipher, used through World War II. Strip ciphers were a stepping stone between manual ciphers and electromechanical cipher machines like Enigma.

Used on: Hitt, Friedman, and Scientific Cryptanalysis

Gathering intelligence from the patterns of communication rather than its content. Even without reading the messages, an analyst can learn a great deal from who is communicating with whom, how often, and when. A sudden increase in radio traffic between a military headquarters and front-line units might signal an upcoming attack, even if the messages themselves are unreadable.

Used on: Black Chamber

The British Admiralty's codebreaking unit during World War I, named after its office number in the Old Admiralty Building. Room 40 was responsible for decoding the Zimmermann Telegram and reading German naval communications throughout the war. Its staff included linguists, classicists, and crossword enthusiasts — foreshadowing the diverse mix of talent at Bletchley Park in World War II.

Used on: Zimmermann Telegram

Military Intelligence, Section 8 — the US Army's codebreaking unit during World War I, created in 1917 under Herbert O. Yardley. MI-8 broke enemy codes and developed secure American code systems. After the war, it evolved into the Cipher Bureau (the Black Chamber).

Used on: Black Chamber

The informal name for the American Cipher Bureau, the US government's peacetime codebreaking organization from 1919 to 1929. Jointly funded by the War Department and State Department, it operated in secret from New York City under Herbert Yardley. Its greatest success was breaking Japanese diplomatic codes before the Washington Naval Conference. It was shut down by Secretary of State Henry Stimson in 1929.

Used on: Black Chamber, Friedman

The US Army's codebreaking organization founded in 1929 under William Friedman, after the Black Chamber was shut down. The SIS brought a more scientific and mathematical approach to cryptanalysis. Friedman's team at SIS broke the Japanese PURPLE cipher machine in 1940. The SIS eventually evolved into the Army Security Agency and then the National Security Agency (NSA).

Used on: Friedman

Intelligence gathered from intercepted communications and electronic signals. SIGINT includes both cryptanalysis (breaking the codes) and traffic analysis (studying the patterns of communication). The organizations discussed on these pages — Room 40, MI-8, the Black Chamber, and the SIS — were all SIGINT organizations.

Used on: Early Modern Cryptanalysis